A reduced order thrust estimation filter is derived, based on a linearized process model and a low order disturbance description. The reduced filter is obtained by ignoring the measurement error on the two spool speeds. Robustness to engine to engine variability is evaluated by Monte Carlo simulations by use of a nonlinear RM12 jet engine model. Both measurement and process disturbance models are presented and used in the filter design. The difference between the reduced order filter and full order filter is negligible. The Monte Carlo simulations clearly show that Kalman based thrust estimation filters are sensitive to parameter variations caused by engine degradation or engine to engine variation.

This content is only available via PDF.
You do not currently have access to this content.